Table 5.

Tests of Heterogeneity among Sites in Polymorphism and Divergence from DNASlider^a

Taxon	Nsequences	Npoly	Nfixed	Runs Test		Kolmogorov–Smirnovb		Mean sliding G
				Test Statisticc	P	Test Statisticd	P	Test statisticd	P
Promoter
Arabidopsis lyrata	23	67	57	56	0.2293	0.0801	0.0334	2.9725	0.2899
Arabidopsis lyrata lyrata	10	40	63	48	0.5293	0.0508	0.3949	2.2286	0.4418
Arabidopsis lyrata petraea	13	49	57	44	0.0846	0.0919	0.0257	3.5192	0.2071
Arabidopsis thaliana	18	15	71	27	0.795	0.0381	0.3685	1.1409	0.557
Gene (silent sites)
A. lyrata	28	123	188	126	0.0122	0.0066	0.0136	6.8013	0.0692
A. l. lyrata	13	69	207	89	0.035	0.0562	0.025	5.6133	0.1015
A. l. petraea	15	99	196	112	0.24	0.0722	0.0089	9.2926	0.0294
A. thaliana	13	37	241	66	0.6776	0.0406	0.0136	5.2926	0.0591

^a For each test, we used the value of the recombination rate parameter (R) that produced the most conservative test statistic.

^b The test using Kolmogorov–Smirnov statistics is most powerful when the ratios at the far ends of the gene differ (McDonald 1998), and this test appears to be the most sensitive to the heterogeneity in our data set.

^c A smaller value indicates greater heterogeneity.

^d A larger value indicates greater heterogeneity.